Developing effective interventions to change or buffer risk due to adolescent friendship networks will require basic research to understand the mechanisms and moderators of network influences. This is a proposal to study the role of friendship networks in risk for HIV transmission among adolescents. The primary focus is on the role that norms of adolescent peer groups play in behaviors and attitudes associated with elevated HIV infection risk. The proposed study will apply knowledge gained in studies of peer networks related to other high-risk behaviors and draw from findings about social networks related to HIV risk of other populations to help increase understanding of the aspects of adolescent peer networks that can mitigate adolescent HIV risk. The specific aims of this proposal are 1) to study the peer network factors that are closely related to risk, including the behavior and beliefs of peers, and the position of the individual with respect to his or her peer network, and 2) to study the buffering or moderating effects of family factors such as family relationships, and parenting practices. These questions will be addressed through secondary analysis of the Add-Health data set. Specifically, the saturated schools subset of the Add-Health data set (n=2654) will be used to create measures of peer networks, average HIV risk behaviors of peers (unprotected sex, drug use, alcohol use), peers' beliefs endorsing unprotected sex, family cohesion, and monitoring across two waves of data collection, using both parent and child sources where possible. Analyses will be longitudinal, and will account for the nesting of individuals within networks and schools through the use of mixed models.